A major portion of the worldwide petrochemical industry is concerned with the production of light olefin materials and their subsequent use in the production of numerous important chemical products via polymerization, oligomerization, alkylation and the like well-known chemical reactions. Light olefins include ethylene, propylene mixtures thereof. These light olefins are essential building blocks for the modern petrochemical and chemical industries. The major source for these materials in present day refining is the steam cracking of petroleum feeds. For various reasons including geographical, economic, political and diminished supply considerations, the art has long sought a source other than petroleum for the massive quantities of raw materials that are needed to supply the demand for these light olefin materials. In other words, the holy grail of the R & D personnel assigned to work in this area is to find a way to effectively and selectively use alternative feedstocks for this light olefin production application, thereby lessening dependence of the petrochemical industry on petroleum feedstocks. A great deal of the prior art's attention has been focused on the possibility of using hydrocarbon oxygenates and more specifically methanol or dimethylether (DME) as a prime source of the necessary alternative feedstock. Oxygenates are particularly attractive because they can be produced from such widely available materials as coal, natural gas, recycled plastics, various carbon waste streams from industry and various products and by-products from the agricultural industry. The art of making methanol and other oxygenates from these types of raw materials is well established and typically involves the use of one or more of the following procedures: (1) manufacture of synthesis gas by any of the known techniques typically using a nickel or cobalt catalyst followed by the well-known methanol synthesis step using relatively high pressure with a copper-based catalyst; (2) selective fermentation of various organic agricultural products and by-products in order to produce oxygenates; or (3) various combinations of these techniques.
Given the established and well-known technologies for producing oxygenates from alternative non-petroleum raw materials, the art has focused on different procedures for catalytically converting oxygenates such as methanol into the desired light olefin products in order to make an oxygenate to olefin (OTO) process. These light olefin products that are produced from non-petroleum based raw materials must of course be available in quantities and purities such that they are interchangeable in downstream processing with the materials that are presently produced using petroleum sources. Although many oxygenates have been discussed in the prior art, the principal focus of the two major routes to produce these desired light olefins has been on methanol conversion technology primarily because of the availability of commercially proven methanol synthesis technology. A review of the prior art has revealed essentially two major techniques that are discussed for conversion of methanol to light olefins (MTO). The first of these MTO processes is based on early German and American work with a catalytic conversion zone containing a zeolitic type of catalyst system. Representative of the early German work is U.S. Pat. No. 4,387,263 which was filed in May of 1982 in the U.S. without a claim for German priority. This '263 patent reports on a series of experiments with methanol conversion techniques using a ZSM-5 type of catalyst system wherein the problem of DME by-product recycle is a major focus of the technology disclosed. Although good yields of ethylene and propylene were reported in this '263 patent, they unfortunately were accompanied by substantial formation of higher aliphatic and aromatic hydrocarbons which the patentees speculated might be useful as an engine fuel and specifically as a gasoline-type of material. In order to limit the amount of this heavier material that is produced, the patentees of the '263 patent propose to limit conversion to less than 80% of the methanol charged to the MTO conversion step. This operation at lower conversion levels necessitated a critical assessment of means for recovering and recycling not only unreacted methanol but also substantial amounts of a DME intermediate product. The focus then of the '263 patent invention was therefore on a DME and methanol scrubbing step utilizing a water solvent in order to efficiently and effectively recapture the light olefin value of the unreacted methanol and of the intermediate reactant DME.
This early MTO work with a zeolitic catalyst system was then followed up by the Mobil Oil Company who also investigated the use of a zeolitic catalyst system like ZSM-5 for purposes of making light olefins. U.S. Pat. No. 4,587,373 is representative of Mobil's early work and it acknowledged and distinguished the German contribution to this zeolitic catalyst based MTO route to light olefins. The inventor of the '373 patent made two significant contributions to this zeolitic MTO route the first of which involved recognition that a commercial plant would have to operate at pressure substantially above the preferred range that the German workers in this field had suggested in order to make the commercial equipment of reasonable size when commercial mass flow rates are desired. The '373 patent recognized that as you move to higher pressure for the zeolitic MTO route in order to control the size of the equipment needed for commercial plant there is a substantial additional loss of DME that was not considered in the German work. This additional loss is caused by dissolution of substantial quantities of DME in the heavy hydrocarbon oil by-product recovered from the liquid hydrocarbon stream withdrawn from the primary separator. The other significant contribution of the '373 patent is manifest from inspection of the flow scheme presented in FIG. 2 which prominently features a portion of the methanol feed being diverted to the DME absorption zone in order to take advantage of the fact that there exist a high affinity between methanol and DME thereby downsizing the size of the scrubbing zone required relative to the scrubbing zone utilizing plain water that was suggested by the earlier German work.
Primarily because of an inability of this zeolitic MTO route to control the amounts of undesired C4+hydrocarbon products produced by the ZSM-5 type of catalyst system, the art soon developed a second MTO conversion technology based on the use of a non-zeolitic molecular sieve catalytic material. This branch of the MTO art is perhaps best illustrated by reference to UOP's extensive work in this area as reported in numerous patents of which U.S. Pat. No. 5,095,163; U.S. Pat. No. 5,126,308 and U.S. Pat. No. 5,191,141 are representative. This second approach to MTO conversion technology was primarily based on using a catalyst system comprising a non-zeolitic molecular sieve, generally a metal aluminophosphate (ELAPO) and more specifically a silicoaluminophosphate molecular sieve (SAPO), with a strong preference for a SAPO species that is known as SAPO-34. This SAPO-34 material was found to have a very high selectivity for light olefins with a methanol feedstock and consequently very low selectivity for the undesired corresponding light paraffins and the heavier materials. This ELAPO catalyzed MTO approach is known to have at least the following advantages relative to the zeolitic catalyst route to light olefins: (1) greater yields of light olefins at equal quantities of methanol converted; (2) capability of direct recovery of polymer grade ethylene and propylene without the necessity of the use of extraordinary physical separation steps to separate ethylene and propylene from their corresponding paraffin analogs; (3) sharply limited production of by-products such as stabilized gasoline; (4) flexibility to adjust the product ethylene-to-propylene weight ratios over the range of 1.5:1 to 0.75:1 by minimal adjustment of the MTO conversion conditions; and (5) significantly less coke make in the MTO conversion zone relative to that experienced with the zeolitic catalyst system.
The classical OTO technology produces a mixture of light olefins primarily ethylene and propylene along with various higher boiling olefins. Although the classical OTO process technology possesses the capability of shifting the major olefin product recovered therefrom from ethylene to propylene by various adjustments of conditions maintained in the reaction zone, the art has long sought an oxygenate to propylene (OTP) technology that would provide better yields of propylene relative to the classical OTO technology. The driving force for this shift in emphasis towards propylene is the growth rate of the propylene market versus the growth rate of the ethylene market. The existing sources of propylene production in the marketplace are primarily based on conventional steam cracking of naphtha, LPG streams, propane streams and the like. Another major existing source of propylene is of course the olefins that are produced in a fluid catalytic cracking (FCC) hydrocarbon conversion process in the modern day refinery. Because the basic raw material used in an OTO process is derived from natural gas which is widely available particularly in remote areas of the world but unfortunately markets for this gas are typically not conveniently available near the location of the remote gas fields. These remote gas fields tend to be quite large and because of the relatively well-developed technology for converting natural gas to methanol and other oxygenates are viewed by those skilled in this art and being the next logical source of large-scale propylene production provided a commercially acceptable OTP process can be established which possesses intrinsic high selectivity for the desired propylene product.
Workers at Lurgi Oel Gas Chemie GmbH have recently announced a new fixed bed methanol to propylene (MTP) process which according to Lurgi offers the potential to satisfy the arts' thirst for a propylene selective OTO type of process. It appears that the basic flow scheme and technical details of the Lurgi process offering in this field have been relatively recently disclosed in a U.S. application publication, Publication No. US2003/0139635A1 which was published on Jul. 24, 2003 and describes a process for selectively producing propylene from a feedstock which comprises methanol and/or DME. Analysis of the two figures attached to this patent publication make it clear that Lurgi contemplates a reactor flow configuration for the alcoholic oxygenate to propylene (OTP) synthesis portion of its flow scheme wherein three reactors are utilized with fixed beds of oxygenate conversion catalysts in a parallel flow arrangement with respect to the oxygenate feed. The reactors are connected in a serial flow arrangement with respect to the effluents of the first reactor and the second reactor. The dual function OTP catalyst system taught as being useful in this flow scheme is rather narrowly defined in this patent publication as a pentasil-type (i.e. ZSM-5 type) having an alkali content less than 380 ppm and a zinc oxide content of less than 0.1 wt-% coupled with a restriction on cadmium oxide content of the same amount. The teachings with respect to this catalyst are derived from a European patent, EP-B-0448000, filed by Sud Chemie and Lurgi claiming priority from an original German application that was filed in March of 1990. Thus the catalyst contemplated for use in Lurgi's flow scheme is well known to those skilled in this art. According to Lurgi's marketing presentation, the on-stream portion of the process cycle for this MTP process is expected to be 500 to 700 hours before in situ regeneration becomes necessary. (See Rothaemel et al. “Demonstrating the New Methanol to Propylene (MTP) Process” presented to the ERTC Petrochemical Conference in March of 2003 at Paris, France). The activity-stability of the OTP catalyst in this Lurgi presentation show a significant drop in conversion activity over the first five cycles with each cycle being terminated after the alcoholic oxygenate conversion level drops to about 94% to 95% of the oxygenate feed. No mention is made in this paper of a corresponding drop in propylene selectivity and instead the average once-through propylene selectivity over the on-stream cycle is discussed and a table presented showing that it ranges from 30% to 40% of the converted products with a number between 68% to 71% presented as an estimate of the average cycle yield for propylene over the 500 to 700 hour cycle length expected to be achieved by this flow scheme with full recycle. Lurgi also contemplates that at the end of the cycle when the conversion has dropped to a level of about 94% of the carbon equivalent in the feed that the reactors will be shut down and the catalyst regenerated using a air/nitrogen mixture to burn off the detrimental coke deposits.
Although Lurgi does not state exactly what countermeasures it takes during its process cycle in order to compensate for the falloff in activity of its dual-function OTP catalyst due to coke deposition, we believe that Lurgi undoubtedly follows the conventional procedure for compensating for activity decay in a catalytic operation involving an increase in the average reactor temperature in order to attempt to hold conversion in the targeted range of greater than 94% of the oxygenate charge. Under these circumstances it is our considered opinion based on experimental results with similar dual-function catalysts and similar feeds that the falloff of propylene selectivity over the cycle is accelerated by the coke deposition and by the attempt to take countermeasures to compensate and we believe that the selectivity falloff will be greater than the activity falloff by a factor of 1.25 to 3.5 or more depending somewhat on the exact composition of the catalyst used and the operating condition changes that are made during the cycle to attempt to compensate for the falloff in activity.
The problem addressed by the present invention is then to modify this OTP process of the prior art in order to enhance its average propylene selectivity over its on-stream cycle time and thereby diminish the requirement for recycle of olefin products other than propylene in order to compensate for lower propylene selectivity. We have now discerned that propylene selectivity is a function not only of reaction conditions but also of average coke level deposited on the OTP conversion catalyst during the on-stream portion of the process cycle. Coke deposition is accelerated when the ΔT (i.e. T at the outlet of the reactor minus T at the inlet) across the individual OTP reaction is not carefully controlled. This is caused by the strongly exothermic nature of the desired OTP reactions which for example at 400 to 500° C. can reach levels of about 7 to 11 kcal/mole of oxygenate converted. This release of exothermic heat of reaction can cause the temperature at the tail-end of the reactor to reach levels that accelerate coke deposition and thus significantly degrade propylene selectivity. Put another way we have now found that average propylene selectivity in an OTP process can be significantly enhanced if the amount of detrimental carbonaceous deposits laid down on the catalyst during the on-stream portion of the process cycle is held to a level such that the OTP catalyst activity is not significantly diminished and oxygenate conversion is maintained over the process cycle near or at essentially start of cycle level. By the use of the term “not significantly diminished” we intend to mean that the conversion level at constant conditions is not allowed to fall more than 2 to 3% during the course of the on-stream catalyst cycle. Quite surprisingly we have also discerned that if this restriction is put on such an OTP process then the average propylene selectivity over the cycle will be improved by about 1.5% to 5.5% or more depending somewhat on the countermeasures taken in the prior art process as the catalyst performance degrades due to the high level of coke deposition on the catalyst during the latter portion of the on-stream cycle. In other words if the once-through propylene selectivity of the prior art OTP process is in the range of 30% to 40% of the carbon-containing products from the OTP reactions we would anticipate the use of our invention to enable an improvement in this range to a level of at least 31.5% to 41.5% or more, thereby significantly improving the economics of the process. In sharp contrast to the OTP process of the prior art, our invention envisions shifting at least 10%, and preferably 20 to 30% or more, of the exothermic heat of reaction liberated when an alcoholic oxygenate such as methanol is converted to propylene from the OTO conversion step to a feed pretreatment step involving an etherification reaction which is designed to convert alcoholic oxygenates such as methanol to their corresponding ether (i.e. DME in the case of methanol) with resulting release of energy that is used to preheat and/or vaporize the feed rather than to accelerate coke deposition in the individual OTP reactors. This strategy allows tighter control of the ΔT across each of the OTP reactors and thus results in less detrimental coke deposition on the OTP catalyst and in lower permanent deactivation due to steam dealumination which accelerates at higher reactor temperatures. When this ΔT control strategy is coupled with our concept of replacing the fixed bed technology in the olefin synthesis reactors of the prior art with moving bed technology, it provides an OTP process that meets the target of not significantly disturbing initial OTP catalyst activity and propylene selectivity while operating with a catalyst circulation rate through the reactors such that the catalyst on-stream cycle time is 300 hours or less. Under these conditions the initial on-stream alcoholic oxygenate conversion of the OTP process does not drop by more than 2 to 3% thereby enabling maintenance of propylene selectivity at essentially start of run condition.
The instant application is a companion case to our prior filed case on the use of moving bed technology in an OTP process filed Jun. 25, 2004 as application Ser. No. 10/876,394, all of the teachings of which are incorporated herein by reference.